Haskell in the Datacentre
simonmar.github.io
simonmar.github.io
Wow, I didn't know this. Anyone knows for which programs they use Haskell? The article doesn't say anything.
This might be it, don't have time to verify right now: http://cufp.org/2015/fighting-spam-with-haskell-at-facebook....
Video: https://www.youtube.com/watch?v=IOEGfnjh92o
Abstract: http://cufp.org/2015/fighting-spam-with-haskell-at-facebook....
I'd be curious to find out if the success FB has had with Haskell for spam filtering means they might consider Haskell for different parts of their stack too? Does anyone have any insight on this?
Ah, yes. As a Go developer, really wish Go moved to an 1:1 threading model.
(I believe the story is slightly different in Haskell.)
Nice sequential blocking-style code is my favorite thing about Go. I realize it's not C, and it makes runtime work much harder, but the payoff up in complex servers is completely worth it.
No, it would not mean such a thing at all. Where did you get this idea?
In fact, on some operating systems/linker combinations, gccgo uses a 1:1 threading model.
> Nice sequential blocking-style code is my favorite thing about Go.
Mine too.
A typical server under load has more outstanding requests to answer than it can OS threads. It is not uncommon in a Go server at Google to find a million or so goroutines.
If you want to have an OS thread that can also run C code, it needs a large enough stack to run C code. A million such stacks is not practical.
On an OS that lets you avoid creating stacks with threads like Linux, and which has quite light-weight kernel-side accounting for threads, it would be possible to give each goroutine a thread. You would still need an N:M pool for cgo.
As an added problem, 1:1 is slower for certain kinds of common channel operations. Anywhere you have an unbuffered channel and block waiting for a value, a 1:1 model requires a context switch to pass the value, whereas M:N means the userland scheduler can switch goroutines on the OS thread.
It is precisely these benchmarks that led Ian to implement M:N in gccgo. If there are combinations where it is 1:1, that is either a new decision he has made in the last year, or (more likely) OS/ARCH combinations that he hasn't moved to M:N yet.
I have seen a similar attempt at 1:1 lightweight tasks in C++ that ran into this. Without the ability to preemptively move tasks between OS threads, it ran into performance problems. Programs that needed the speed in that model had to switch to a futures-style of programming.
Yes, I've been wanting to write a proposal for over 5 years now, if only there was more time in the world...
Coincidentally, I also have a method for implementing Go stacks (growing and shrinking) that does not involve copying stacks and avoids the hot-split problem.
Late edit:
In fact, this method would work particularly well for "growing" the stack to call C code in the 1:1 model proposed above.
It would make calling a C function, just another function call with some marker on it (like we have NOFRAME and NOSPLIT now). And when we switch to a more ABI-compatible calling convention (which we want to do anyway, irrespective of all this), then we would not even be a need for a C translator written in assembly any more.
Quite literally we would get cgo for free.
First of all, I object to the use of "can" here. There was a time where this was true, but, as I said, that time has passed. There is no problem running millions of kernel threads today.
That being said, there are advantages of not doing that. Right now Go uses a network polling mechanism like epoll to implement network I/O without consuming threads. This is totally orthogonal to the 1:1 vs. N:M discussion. Go can use 1:1 scheduling while still performing asynchronous I/O behind the scenes, just as it does today.
The nature of the mapping between g's and m's does not influence the other {g, g, ...} -> {g0{m}} mapping used to implement network I/O.
> It is not uncommon in a Go server at Google to find a million or so goroutines.
No problem with that, they can continue to do so.
> If you want to have an OS thread that can also run C code, it needs a large enough stack to run C code. A million such stacks is not practical.
C code needs its own stack, but this is again orthogonal to scheduling. Right now there is a N:1 mapping {g, g, ...} -> {g0} which makes it easy to switch to g0 stack. This will have to change to a N:M mapping {g, g, ...} -> {g0, g0, ...} and Go code would have to acquire C stacks, just as it acquires any other resource that it needs to use.
This is not expensive, in fact, C calls are now expensive precisely because there is a deep interaction between calling C code and the scheduler (that needs thousands of cycles). All this cycles will Go away. In a 1:1 model all you'd have to do is acquire a C stack, which is very fast and happens in userspace for the uncontented case (the common case), probably less than a hundred cycles.
> You would still need an N:M pool for cgo.
As explained above, you need a N:M pool, but this does not need to be integrated with the scheduler any more, making it much simpler to implement (and more performant).
> Anywhere you have an unbuffered channel and block waiting for a value, a 1:1 model requires a context switch to pass the value, whereas M:N means the userland scheduler can switch goroutines on the OS thread.
Only if you naively implement channels as mutex-based queues.
There is still a runtime underneath that can switch stacks on different threads. I want to move general scheduling decisions out of the Go runtime and into the kernel, but pointwise, stateless, swap-of-control type of things can still happen synchronously in the Go runtime.
I object to the use of "can", just because we can doesn't mean we should.
Last I checked, creating a thread (on a Windows) took 16 ms. Creating 1 million threads would take more than 4 hours. That's the hell of a startup time! =D
(And don't bother saying Linux is faster. It's not relevant unless it's by 4 orders of magnitude ;) ).
But as I said I'm happy to be convinced. The easiest way to demonstrate it would be to move gccgo linux/amd64 back to 1:1 without hurting the channel benchmarks. You could use that as an argument that fanning out an epoll loop among threads can be made fast.
Ok i am a bit tongue in cheek, but still.
Haskell would lose more features, but also its "blocking-style" syntax (that is based on callbacks, but nobody ever thinks about it this way).
There is a reason POSIX threading is implemented in the kernel on every operating system, even if they all started as N:M or 1:M implementations.
What could you be hoping to accomplish?
Because it's already there and it already works just fine.
It's so frustrating to see this sort of comment on nearly every post about a language outside the mainstream. It's a religious argument and not a productive one.
Well clearly it didn't work just fine if its threading model was interacting poorly with that of their Haskell code. If there's a mismatch between those two components then that will continue to bite them in the future, and it's legitimate to ask whether moving the boundary would be a better way to solve the problem.
When they realized the bug the path forward probably seemed like patching and not a Haskell re-write, and that's a reasonable mistake to make.
They seemed to have succeeded in creating a culture of coders/hackers that were opinionated/empowered sufficiently to use their tool of choice, and to persuade others of the advantages.
This doesn't really happen in large-cap companies (though does happen to empowered teams, but even then, only somewhat - thinking finance). And speaks to advantages.